CN106251016A - A kind of parking system paths planning method based on dynamic time windows - Google Patents

Abstract

The invention discloses a kind of parking system paths planning method based on dynamic time windows, belong to Path Planning Technique field, it is characterised in that comprise the following steps: step S1: use topological approach to create the working environment model of AGV in intelligent garage；Step S2: according to different evaluation standard, respectively every AGV and each Transport Vehicle task setting priority；Step S3: using dijkstra's algorithm is that the AGV received an assignment plans the shortest feasible path；Step S4: arrangement feasible path time window；Step S5: different according to conflict type, design conflicts resolution policy；Step S6: utilize parking system path planning algorithm based on dynamic time windows to plan Lothrus apterus optimal path for AGV.The present invention uses timesharing Utilization strategies, by dijkstra's algorithm and time window method are effectively combined, it is possible not only to effectively to solve that current many AGV path planning flexibility is poor, easily the problem such as deadlock, collision conflict occur, and can be the AGV the received an assignment Lothrus apterus path optimizing of cooking up a shortest time.In addition, the present invention also can be effectively improved intelligent three-dimensional shutdown system overall operation efficiency, reduces society personnel and deposits, withdraws the car waiting time.

Description

A kind of parking system paths planning method based on dynamic time windows

Technical field

The invention belongs to automatical pilot transportation vehicle (Automatic guided vehicle is called for short AGV) path planning skill Art field, is specifically related to a kind of parking system paths planning method based on dynamic time windows.

Background technology

In recent years, along with the fast development of national economy, China's automobile pollution sharply increases.According to Traffic Administration Bureau of the Ministry of Public Security Announcement data show, by the end of in by the end of December, 2015, China's vehicle guaranteeding organic quantity reaches 2.79 hundred million, with increasing about 1500 compared with last year Ten thousand, wherein automobile pollution is 1.72 hundred million, accounts for the 61.6% of sum.In China, automobile pollution surpasses million City have 40, wherein Beijing, Chengdu, Shenzhen, Tianjin, Shanghai, Suzhou, Chongqing, Guangzhou, Hangzhou, Zhengzhou and Xi'an etc. The automobile pollution in 11 cities is more than 2,000,000.Sharply increasing of automobile pollution causes urban traffic congestion, stops and be stranded Nan Deng social problem, has had a strong impact on civic living environment, and therefore, solving parking difficulty has become the difficulty that society is urgently to be resolved hurrily Topic.And the appearance of horizontal mobile intelligent stereo garage of based on AGV, solve this difficult problem the most well.This Intelligent standing Body garage parking is similar to automatic stored storage device, is realized by equipment collaboration effects such as AGV, elevator and trailer plates Park function with layer or different layers vehicle access, have that parking floor space bicycle parking quantity few, effective is many, vehicle access automatization Degree is high, cost performance is high and security reliability advantages of higher, can realize unmanned automatically storing and taking vehicles, AGV automatic charging and car The various functions such as storehouse automatic charging.In this intelligent stereo garage, owing to its running environment is complicated and changeable, how to make in system AGV effectively avoid path resources competition and conflict on the premise of, the access smoothly completing vehicle in the short period of time stops Letting alone business, this relates to many AGV collaborative obstacle avoidance path planning problem.

Path planning is the important step of AGV airmanship, and it refers in the environment with barrier, according to certain Evaluation criterion (such as beeline, the least cost time, minimum number of turns and minimum energy resource consumption etc.), find one from Path is touched to the optimum of target location or the nothing of near-optimization in beginning position.

Due to separate unit AGV working capability finite, it is difficult to complete complex task, therefore, in intelligent stereo garage, need Multiple stage AGV jointly completes vehicle access and parks task.Multiple stage AGV path planning is different from separate unit AGV path planning, separate unit AGV The essence of path planning is route searching problem, i.e. searches out the starting point path to impact point in a map, and makes A certain performance indications are optimum.Many AGV path planning is then more complicated than single AGV path planning many, at operation ring complicated and changeable Under border, its separate unit AGV to be searches out a path optimizing from starting point to impact point, and AGV to be avoided and around Collision is clashed between Environment Obstacles thing and other AGV.In addition, the coordination between its multiple stage AGV to be completed, it is to avoid Collision, Deadlock occur, in order to make to smoothly complete appointed task by synergism between multiple stage AGV.

For many AGV path planning problem, the most relevant scholar has done numerous studies work, and in succession proposes many Plant effective ways, such as fuzzy inference system method, Petri network method, hybrid multi-objective genetic algorithm, distributed control methodology, time window method And the paths planning method etc. that time window is combined with other heuritic approaches.Although said method can solve many AGV path rule The problem of drawing, but it there is also many defects, as algorithm complicated calculations amount is big, system whole efficiency is low, be prone to generation deadlock and resistance Fill in, be difficult to obtain globally optimal solution and environmental suitability and poor universality etc..Deposit for solving many AGV in intelligent stereo garage Pick up the car path planning problem, it is to avoid route searching occurs deadlock and collision conflict, improves robustness and the flexibility of existing algorithm, Improving intelligent three-dimensional shutdown system overall operation efficiency, reduce society personnel and deposit, withdraw the car waiting time, the present invention proposes one Parking system paths planning method based on dynamic time windows.

The one of the propositions such as a kind of based on dynamic time windows the parking system paths planning method of present invention proposition and Qiao Hui Plant multi-robots Path Planning Method based on time window and there is the difference of essence.Both identical points are all at dijkstra's algorithm With propose on the basis of time window method, and be used for this blending algorithm solving producing practical problem.Difference is: 1. excellent First level aspect, the present invention is the shortest as evaluation criterion with car number size, task loading sequence, the task order of importance and emergency and distance, It is respectively every AGV and each Transport Vehicle task setting priority.It addition, for crossing collision problem, supplement again and be provided with Distance priority level；And in the method that Qiao Hui provides, do not provide priority is specifically related to object range and priority evaluation Standard；2. in terms of time window, the present invention arrange feasible path time window then use time window initialize, time window update and Time window arranges three steps, updates for time window and is then mainly used in checking between different task the time window of feasible path whether There is overlapping phenomenon to be loaded onto in time window vector table with by the time window removed or unoccupied section is corresponding, time window The calculating process of arrangement program is the most relevant with loading tasks；And in the method that Qiao Hui provides, arrangement feasible path time window is then Only using time window to update a step to complete, and the circulation renewal time of time window is fixing, it updates the mesh of time window Determine that feasible path time window arrangement；3. collision detection aspect, the present invention is the distance between comprehensive AGV, runs section Length and the factor such as the AGV speed of service and traffic direction, solve collision detection problem；And the method provided at Qiao Hui In, it is by judging whether all directed edges form ring, solves collision detection problem；4. in conflict-solving strategy side Face, the present invention is different according to conflict type, devise deceleration strategies, waiting strategy and again path planning strategy (this strategy is again Including local paths planning strategy and global path planning strategy), as intersection conflict and in opposite directions conflict in keep away Exempt from conflict, use waiting strategy to solve；For the inevitable conflict in conflict in opposite directions, use path planning strategy solution again Certainly；For catching up with and surpassing conflict, use deceleration and waiting strategy to solve, it is possible to according to actual needs, use local paths planning plan Slightly；And in the method that Qiao Hui provides, the robot solved for nothing in deadlock, it is only to record rushing on its every feasible path Prominent short path, conflicting nodes and conflict dependence, provide early warning to being relied on most robots；5. process is performed at algorithm Aspect, both have essence difference, this difference can be can be seen that by the path planning process figure in two kinds of methods.

Summary of the invention

Present invention aim at for AGV Transport Vehicle path planning problem in intelligent stereo garage, many taking into full account Under AGV collaborative obstacle avoidance and environment time variation precondition, use timesharing Utilization strategies, by by dijkstra's algorithm and time Window method effectively combines, it is provided that a kind of parking system paths planning method based on dynamic time windows.The present invention can effectively keep away Exempt from deadlock and collision problem, it is ensured that path planning is optimum, and has preferable flexibility under dynamic environment.

The present invention is achieved through the following technical solutions its technical purpose, a kind of parking system path based on dynamic time windows Planing method, comprises the steps:

Step S1: use topological approach to create the working environment model of AGV in intelligent garage；

Step S2: according to different evaluation standard, respectively every AGV and each Transport Vehicle task setting priority；

Step S3: using dijkstra's algorithm is that the AGV received an assignment plans the shortest feasible path；

Step S4: arrangement feasible path time window；

Step S5: different according to conflict type, design conflicts resolution policy；

Step S6: utilize parking system path planning algorithm based on dynamic time windows to plan Lothrus apterus optimum road for AGV Footpath；

Further, in described step S1, topological approach is used to create the working environment model of AGV in intelligent garage, specifically Step includes:

Step S11: the traffic network in environmental model and AGV are handled as follows: 1. AGV runs track is that single track is two-way Pattern, and width is only capable of accommodating an AGV；2. the AGV in system can only accept a Transport Vehicle within the same time period Task, during its execution task, it is invalid that other tasks of system distribution are then considered as；3. for avoiding colliding with other AGV Accident, the AGV being required to be execution task sets a safety traffic region, and this safety zone can be by AGV car body physical dimension, operation Speed and operation track physical dimension determine；4. the arbitrary intersection within certain moment or certain time period, in road network The most only allow an AGV to use with arbitrary running section；

Step S12: utilize photographic head, radar sensor, ultrasonic sensor and infrared ray sensor etc. that AGV carries Gathering AGV running environment information, above-mentioned information includes the initial parking stall of AGV, target parking stall, barrier and AGV position to be charged Put；

Step S13: create AGV in intelligence as modeling data, employing topological approach using the environmental information that aforesaid operations gathers Working environment model in garage.

Further, in described step S2, according to different evaluation standard, respectively every AGV and each Transport Vehicle task Setting priority, particular content includes:

Step S21: for the priority of AGV in system, then determined by car number size, and AGV priority height low order Sequence becomes negative correlation with car number size；

Step S22: for the priority of Transport Vehicle task in system, then by task loading sequence, the task order of importance and emergency and The evaluation criterions such as distance is the shortest comprehensively determine；

Step S23: when there is intersection conflict, for the AGV sequencing problem by intersection of conflicting, then Comprehensively determined by AGV current priority and the shortest priority of distance；

Step S24: system it further provides for, the priority of the AGV being carrying out Transport Vehicle task is higher than the preferential of idle AGV Level；During AGV execution task, ground control system is that the new Transport Vehicle task of its distribution is considered invalid.

Further, in described step S23, when intersection conflict occurs, for AGV by conflict intersection Sequencing problem, then situation about comprehensively being determined by AGV current priority and the shortest priority of distance includes:

Step 231: when two AGV arrive same intersection simultaneously, first AGV priority judged by system, Then according to priority height order, arrange two AGV sequencing by intersection.The AGV high when priority leads to After crossing the safe distance that intersection and the AGV low with priority keep certain, system can continue to hold by the low AGV of call priority Row task；

Step 232: when two AGV are one in front and one in back to arrive intersection, but when both meetings conflict occur in intersection, Now system is on the basis of judging AGV priority, also to determine each other according to the length of two AGV to intersection distance By the sequencing at crossing；

Step 233: when two AGV that priority is identical arrive intersection simultaneously, system can be according to two AGV distances The distance of intersection determines its sequencing by intersection.

Further, in described step S3, using dijkstra's algorithm is that the AGV received an assignment plans the shortest feasible path It is critical that its must according in step S2 priority height order carry out.For task m any one of system_iLetter Number may be defined as:

m_i(t)=(s_i,d_i,λ_i(t),P_i(t),q_i)

In formula, i represents mission number；m_iT () represents the task of t system distribution；s_iExpression task m_iStarting point, d_i Expression task m_iTerminal, λ_iExpression task m_iThe set in a series of orderly section of process, P_iT () represents task m_iPreferential Level, q_iRepresent execution task m_iAGV.After many AGV path planning terminates, the above-mentioned parameter of each task typically immobilizes, The most in case of a collision, the AGV that some priority is low just needs dynamically to change its running route, avoids holding with this Collide between the AGV of row task, deadlock conflict and strengthen AGV flexibility.

Further, in described step S4, feasible path time window of arranging, concrete steps include:

Step S41: time window initializes.After the shortest feasible path determines, under ideal conditions (Lothrus apterus), for accepting to appoint The AGV of business arranges out feasible path time window.By task m in step S3_iThe shortest feasible path λ found out_i, it is by a series of Operation section forms, available orderly section set expression, i.e. λ_i={ e_j,e_k,e_l,…,e_q, e_j,e_k,e_l,…,e_q∈ E, wherein, E represents the set in all feasible sections, e in road network_k(k ∈ 1,2,3 ...., q) represent a certain section in the shortest feasible path.

Task m_iAt section e_kOn time window function may be defined as:

T_w,ik=(q_i,m_i,r,t_in,k,t_out,k)

In formula, r represents section e_kAt feasible path λ_iOn position；t_in,kRepresent vehicle q_iSail section e into_kInitial time Between；t_out,kRepresent vehicle q_iLeave section e_kTime.

For section e_kTime window, can be calculated by following formula:

t_out,k=t_in,k+ω_i,k

In formula, ω_i,kRepresent that AGV is by section e_kThe time spent, can be calculated by following formula:

${\mathrm{\ω}}_{i,k}=\frac{l_{i,k}}{v}$

In formula, l_i,kRepresent section e_kLength, v represents the speed of service of AGV.

In actual applications, due to the section e the most in order of feasible path_kNeed to be used by AGV timesharing, therefore, in order Section e_kAlso it is made up of a series of time windows, available ordered vector Represent.In order to AmountIn, vector dimension is identical with Transport Vehicle task quantity, can change over and change.If task m_iDo not have in certain moment Use section e_k, then can be time of sailing into the t in this section_in,kWith roll time t away from_out,kIt is both configured to 0.Further, since task m_i The shortest feasible path be made up of a series of orderly sections, and every orderly section correspond to a time window, therefore, appoints Business m_iIt is also believed to be made up of a series of time windows, usable setRepresent.

According to equation given in step S41, it can be task m_iThe shortest feasible path λ_iArrange out such as set D_iShown Time window is distributed.

Step S42: time window updates.Arrange out behind a time window path ideally according to step S41, then Check between different task, whether the time window of feasible path exists overlapping phenomenon.

If non-overlapping phenomenon, then task m_iPath planning process terminate.If current task m_iIt is preferential in current system During the highest scheduler task of level, then feasible path time window step S41 planned is as task m_iFinal time window, it is not necessary to Again update.

If there being lap, then on explanation current task and the shortest feasible path that goes out of other mission plannings at least one Section uses simultaneously.For this kind of phenomenon, then need system according to conflict type, conflict Robot dodge strategy reasonable in design.

Step S43: the arrangement of time window.By Lothrus apterus time window corresponding for each for feasible path section in a certain order Arrangement, i.e. completes the time window arrangement of feasible path.If it should be noted that a certain section e_kThere is the time of multiple task Window, then the task that is newly added is at section e_kOn the entry time of time window must be fulfilled for condition: the time 1. sailing this section into must Must be more than or equal to the AGV time departure from a upper section；2. the length of the free time window in this section should be greater than or is equal to The time that AGV is spent by this section.

According to step S3 and step S4, searched for by loop iteration, the AGV received an assignment can be followed successively by and cook up Lothrus apterus The shortest feasible path time window.

Further, in described step S4, time window refers to that the AGV performing Transport Vehicle task leaves certain from initially entering to The time that the whole process in individual intersection or certain section is spent, its Main Function is the intersection taken AGV Or running section is marked, to avoid within the time period that this AGV takies, is used by other AGV and cause deadlock or collision Conflict.

Further, in described step S4, the position that ground control system meeting real-time reception AGV uploads during exercising Put, the information such as speed and running status, and according to these information, it is judged that whether AGV has been moved off certain section or certain intersects Crossing, drives towards next section or intersection.If AGV has been moved off certain section or certain intersection, then need to be from the time Window vector table is deleted the information that AGV registers in the table, thus discharges this section or this intersection resource, for other AGV Use.

Further, in described step S5, different according to conflict type, design conflicts resolution policy, particular content includes:

The conflict occurred in many AGV Transport Vehicle path planning typically has two kinds, i.e. intersection conflict and path conflict.

Step S51: intersection conflict refers to be carved with two or more AGV because sharing a friendship simultaneously in the some time Cross road mouth and the conflict that causes.Conflicting for such, system typically uses waiting strategy to solve, i.e. system is low by priority The time window that uses of AGV application translate a period of time backward, until the high AGV of priority passes through conflict intersection After, then this intersection of application use, avoid deadlock and collision conflict with this.When multiple stage AGV arrives a certain intersection simultaneously Time, first each AGV priority judged by system, then according to priority height sequencing AGV is by intersection first Rear order.

Step S52: path conflict can be divided into again and conflicts in opposite directions and catch up with and surpass conflict, wherein, conflicts in opposite directions and can be divided into and can keep away Exempt from conflict and inevitable conflict.For may wait for avoiding conflict, system can use waiting strategy to solve；For inevitable punching Prominent, path planning strategy again can be used, this strategy includes local paths planning strategy and global path planning strategy；For catching up with Overshoot is dashed forward, and deceleration and waiting strategy can be used to solve, it is possible to according to actual needs, uses local paths planning strategy to feasible The conflict section in path carries out planning again and processes.

Further, in described step S52, in conflict refers to certain time period in opposite directions, same path is transported in opposite directions The conflict caused for contention path resources between the AGV of row.Catch up with and surpass conflict refer to two AGV run the most on one path and The speed of service of the rear side AGV speed of service higher than front side AGV, the conflict caused for contention path resources between them.

Further, in described step S52, local paths planning strategy refers in not utilizing garage traffic network whole Road section information and in not changing the shortest feasible path on the premise of Lothrus apterus section, only on other roads adjacent with conflict section Duan Zhong, searches out other sections that can substitute conflict section, and this section can guarantee that and do not affecting on the shortest feasible path On the premise of the distribution of other times window, can guarantee that the AGV received an assignment arrives at the destination smoothly, the Transport Vehicle completing to specify is appointed Business.

Further, in described step S52, global path planning strategy refers to utilize whole sections in garage traffic network Information, the AGV for receiving an assignment cooks up the shortest feasible path again.

Further, in described step S52, the step that is embodied as of path planning includes again:

Step S521: system detects the inevitable conflict whether occurring in conflicting in opposite directions between many AGV；

Step S522: the section inevitable conflict occur is marked, and calls path search algorithm and again plan Path；

Step S523: to new search to the time window vector table of path optimizing be circulated renewal, until cooking up nothing Till the path optimizing that conflict is minimum with the time, algorithm search terminates.If algorithm is searched for (for avoiding program through successive ignition Endless loop occur, program cyclic search number of times is provided with maximum restriction) still cannot find Lothrus apterus path optimizing, then algorithm search Process terminates, and this task is loaded in task sequence table, and waiting system task scheduling next time is distributed.

Further, in described step S6, parking system path planning algorithm based on dynamic time windows is utilized to advise for AGV Drawing Lothrus apterus optimal path, concrete steps include:

Step S61: each parameter of initialization algorithm, sets up set N, set Q, set M, set A, set S and set R, point Not Yong Yu AGV in storage system, the AGV received an assignment, the instruction of Transport Vehicle task requests, priority policy process after access Car task and the starting point of task and impact point；

Step S62: the Transport Vehicle request instruction in set M is loaded in task sequence set A, and according to priority of task Level height order is its sequence；

Step S63: judge AGV duty according to AGV current state information, such as idle condition and be carrying out task shape State, state can represent with 0 and 1.If available free AGV exists in working environment, then system can be by preferential in task sequence set A The highest Transport Vehicle task of level distributes to the idle AGV that numbering is minimum, and in known AGV start position, aiming spot and work On the premise of making the information such as environment, call dijkstra's algorithm, cook up a shortest path optimizing of distance for AGV, then Go to step S64；Otherwise, then task scheduling stops；

Step S64: calculate AGV time of sailing in each section and roll the time away from path optimizing, initial according to step S41 Change each section time window vector table, by repeatedly cyclic search, the time window distribution of the shortest i.e. available feasible path, then turn To step S65；

Step S65: the path of suboptimum task is planned, goes to step S63, it is judged that in environment, no available free AGV can Calling, if having, then the idle AGV that numbering is first only second to priority the highest by system distributes to suboptimum task, then calls Dijkstra's algorithm is that suboptimum mission planning goes out a path optimizing；Otherwise, then the scheduling to this task is suspended, then according to appoint Other tasks are scheduling by business sequence table order successively；If nothing, then system stops dispatching follow-up work；

Step S66: calculate AGV time of sailing in each section and roll the time away from sub-optimal path, update each section time Window vector table, then judges whether time window vector table exists overlap.If time window is non-overlapping, then the path rule of suboptimum task Streak journey to terminate；Otherwise, then need which kind of conflict the path optimizing detecting suboptimum task exists, and according to conflicting type not With, select suitable conflict-solving strategy, as intersection conflict and in opposite directions conflict in the conflict avoided, can use Treat strategy solution；For the inevitable conflict in conflict in opposite directions, path planning strategy can be used again to solve；For catching up with and surpassing punching Dash forward and deceleration and waiting strategy can be used to solve, it is possible to according to actual needs, use local paths planning strategy to feasible path Conflict section carry out planning again process；

Step S67: after the path planning of suboptimum task terminates, go to step S63, then according to task sequence tabular order sequence weight Multiple aforesaid operations, is sequentially completed the path planning of other tasks.

Further, in described step S63, system is the letter uploaded in real time with AGV self to the judgement of AGV status information Breath is as foundation, and the idle condition of AGV represents with 0, is carrying out task status and represents with 1.

Further, in described step S63, before the shortest feasible path of the mission planning that system is distributed, need to first judge Whether system exists idle AGV.If there is idle AGV, system just can be to mission planning feasible path；Otherwise, system cannot For the shortest feasible path of mission planning.

Further, in described step S63, feasible path the shortest to task planning with to accepting the shortest of this task AGV Feasible path planning is the same.

Further, in described step S65, suspended task is dispatched in two kinds of situation, respectively: 1. system is without idle AGV Schedulable uses, then system can stop dispatching follow-up work；2. in system, available free AGV exists, but system cannot be for this Lothrus apterus feasible path is cooked up in business.Now, system only suspends the scheduling to this task, and it has no effect on other tasks Scheduling.

Further, in described step S66, update each section time window vector table, it is judged that each section time window window vector Whether table exists overlapping phenomenon can operate according to step S42 and step S5.

Further, in described step S66, the time window arrangement of the shortest feasible path of suboptimum task can be counted according to step S4 Obtain.

Further, in described step S66, time window overlap problem, the strategy solution that can provide according to step S5.

Beneficial effect

(1) can effectively solve that current many AGV path planning flexibility is poor, easily deadlock, the collision problem such as conflict occur；

(2) can be on the premise of effectively solving path conflict, the AGV for receiving an assignment cooks up the optimization road of shortest time Footpath；

(3) automatic access of the automated management and vehicle that contribute to realizing parking apparatus is parked, and is of value to enhancing system Safety and reliability, improve garage parking apparatus and the utilization rate of parking position, reduce human cost, operation cost and equipment Cost etc.；

(4) intelligent three-dimensional shutdown system overall operation efficiency can be effectively improved, when reduction society personnel deposit, withdraw car wait Between.

Accompanying drawing explanation

Fig. 1 is many AGV based on time window path planning algorithm flow chart；

Fig. 2 is the working environment model of AGV in certain moment intelligent garage；

Fig. 3 is intersection conflict；

Fig. 4 is that waiting strategy solves intersection time window conflict；

Fig. 5 is for conflict in opposite directions；

Fig. 6 is that waiting strategy solves path conflict；

Fig. 7 is for catching up with and surpassing conflict；

Detailed description of the invention

Below in conjunction with accompanying drawing, present invention is described in detail, but is not limitation of the invention.

The present invention is to provide a kind of parking system paths planning method based on dynamic time windows, Fig. 1 show this Bright algorithm implementing procedure figure, the flow chart describes the solution procedure of many AGV Lothrus apterus optimal path, specifically includes following step Rapid:

Step S1: using topological approach to create the working environment model of AGV in intelligent garage, concrete steps include:

Step S11: the traffic network in environmental model and AGV are handled as follows: 1. AGV runs track is that single track is two-way Pattern, and width is only capable of accommodating an AGV；2. the AGV in system can only accept a Transport Vehicle within the same time period Task, during its execution task, it is invalid that other tasks of system distribution are then considered as；3. for avoiding colliding with other AGV Accident, the AGV being required to be execution task sets a safety traffic region, and this safety zone can be by AGV car body physical dimension, operation Speed and operation track physical dimension determine；4. the arbitrary intersection within certain moment or certain time period, in road network The most only allow an AGV to use with arbitrary running section；

Step S12: utilize photographic head, radar sensor, ultrasonic sensor and infrared ray sensor etc. that AGV carries Gathering AGV running environment information, above-mentioned information includes the initial parking stall of AGV, target parking stall, barrier and AGV position to be charged Put；

Step S13: create AGV in intelligence as modeling data, employing topological approach using the environmental information that aforesaid operations gathers Working environment model in garage.

Fig. 2 show AGV working environment illustraton of model in certain the moment intelligent garage using topological approach to create, dark circles in figure Lattice represent and take parking position, and white circular lattice represent idle parking position, and P0 represents garage port, PE (overlap with crossing C2 passage, Do not mark herein) represent garage exit, P1～P15 is garage parking position, can be used for storage of cars, C1～C6 represents that track intersects Crossing, AGV can complete herein to turn to and switch track operation.

Step S2: according to different evaluation standard, respectively every AGV and each Transport Vehicle task setting priority, specifically Content includes:

Step S21: for the priority of AGV in system, then determined by car number size, and AGV priority height low order Sequence becomes negative correlation with car number size；

Step S22: for the priority of Transport Vehicle task in system, then by task loading sequence, the task order of importance and emergency and The evaluation criterions such as distance is the shortest comprehensively determine；

Step S23: when there is intersection conflict, for the AGV sequencing problem by intersection of conflicting, then Comprehensively determined by AGV current priority and the shortest priority of distance；

Step S24: system it further provides for, the priority of the AGV being carrying out Transport Vehicle task is higher than the preferential of idle AGV Level；During AGV execution task, ground control system is that the new Transport Vehicle task of its distribution is considered invalid.

Further, in described step S23, when intersection conflict occurs, for AGV by conflict intersection Sequencing problem, then situation about comprehensively being determined by AGV current priority and the shortest priority of distance includes:

Step 231: when two AGV arrive same intersection simultaneously, first AGV priority judged by system, Then according to priority height order, arrange two AGV sequencing by intersection.The AGV high when priority leads to After crossing the safe distance that intersection and the AGV low with priority keep certain, system can continue to hold by the low AGV of call priority Row task；

Step 232: when two AGV are one in front and one in back to arrive intersection, but when both meetings conflict occur in intersection, Now system is on the basis of judging AGV priority, also to determine each other according to the length of two AGV to intersection distance By the sequencing at crossing；

Step 233: when two AGV that priority is identical arrive intersection simultaneously, system can be according to two AGV distances The distance of intersection determines its sequencing by intersection.

Step S3: use dijkstra's algorithm be the AGV that receives an assignment plan the shortest feasible path it is critical that it must Must carry out according to the priority height order in step S2.For task m any one of system_iFunction may be defined as:

m_i(t)=(s_i,d_i,λ_i(t),P_i(t),q_i)

In formula, i represents mission number；m_iT () represents the task of t system distribution；s_iExpression task m_iStarting point, d_i Expression task m_iTerminal, λ_iExpression task m_iThe set in a series of orderly section of process, P_iT () represents task m_iPreferential Level, q_iRepresent execution task m_iAGV.After many AGV path planning terminates, the above-mentioned parameter of each task typically immobilizes, The most in case of a collision, the AGV that some priority is low just needs dynamically to change its running route, avoids holding with this Collide between the AGV of row task, deadlock conflict.

Step S4: arrangement feasible path time window, concrete steps include:

Step S41: time window initializes.After the shortest feasible path determines, under ideal conditions (Lothrus apterus), for accepting to appoint The AGV of business arranges out feasible path time window.By task m in step S3_iThe shortest feasible path λ found out_i, it is by a series of Operation section forms, available orderly section set expression, i.e. λ_i={ e_j,e_k,e_l,…,e_q, e_j,e_k,e_l,…,e_q∈ E, wherein, E represents the set in all feasible sections, e in road network_k(k ∈ 1,2,3 ...., q) represent a certain section in the shortest feasible path.

Task m_iAt section e_kOn time window function may be defined as:

T_w,ik=(q_i,m_i,r,t_in,k,t_out,k)

In formula, r represents section e_kAt feasible path λ_iOn position；t_in,kRepresent vehicle q_iSail section e into_kInitial time Between；t_out,kRepresent vehicle q_iLeave section e_kTime.

For section e_kTime window, can be calculated by following formula:

t_out,k=t_in,k+ω_i,k

In formula, ω_i,kRepresent that AGV is by section e_kThe time spent, can be calculated by following formula:

${\mathrm{\ω}}_{i,k}=\frac{l_{i,k}}{v}$

In formula, l_i,kRepresent section e_kLength, v represents the speed of service of AGV.

In actual applications, due to the section e the most in order of feasible path_kNeed to be used by AGV timesharing, therefore, in order Section e_kAlso it is made up of a series of time windows, available ordered vector Represent.In order to AmountIn, vector dimension is identical with Transport Vehicle task quantity, can change over and change.If task m_iDo not have in certain moment Use section e_k, then can be time of sailing into the t in this section_in,kWith roll time t away from_out,kIt is both configured to 0.Further, since task m_i The shortest feasible path be made up of a series of orderly sections, and every orderly section correspond to a time window, therefore, appoints Business m_iIt is also believed to be made up of a series of time windows, usable setRepresent.

According to equation given in step S41, it can be task m_iThe shortest feasible path λ_iArrange out such as set D_iShown Time window is distributed.

Step S42: time window updates.Arrange out behind a time window path ideally according to step S41, then Check between different task, whether the time window of feasible path exists overlapping phenomenon.

If non-overlapping phenomenon, then task m_iPath planning process terminate.If current task m_iIt is preferential in current system During the highest scheduler task of level, then feasible path time window step S41 planned is as task m_iFinal time window, it is not necessary to Again update.

If there being lap, then on explanation current task and the shortest feasible path that goes out of other mission plannings at least one Section uses simultaneously.For this kind of phenomenon, then need system according to conflict type, conflict Robot dodge strategy reasonable in design.

Step S43: the arrangement of time window.By Lothrus apterus time window corresponding for each for feasible path section in a certain order Arrangement, i.e. completes the time window arrangement of feasible path.If it should be noted that a certain section e_kThere is the time of multiple task Window, then the task that is newly added is at section e_kOn the entry time of time window must be fulfilled for condition: the time 1. sailing this section into must Must be more than or equal to the AGV time departure from a upper section；2. the length of the free time window in this section should be greater than or is equal to The time that AGV is spent by this section.

According to step S3 and step S4, searched for by loop iteration, the AGV received an assignment can be followed successively by and cook up Lothrus apterus The shortest feasible path time window.

Further, in described step S4, time window refers to that the AGV performing Transport Vehicle task leaves certain from initially entering to The time that the whole process in individual intersection or certain section is spent, its Main Function is the intersection taken AGV Or running section is marked, to avoid within the time period that this AGV takies, is used by other AGV and cause deadlock or collision Conflict.

Further, in described step S4, the position that ground control system meeting real-time reception AGV uploads during exercising Put, the information such as speed and running status, and according to these information, it is judged that whether AGV has been moved off certain section or certain intersects Crossing, drives towards next section or intersection.If AGV has been moved off certain section or certain intersection, then need to be from the time Window vector table is deleted the information that AGV registers in the table, thus discharges this section or this intersection resource, for other AGV Use.

Step S5: different according to conflict type, design conflicts resolution policy, particular content includes:

Step S51: intersection conflict refers to be carved with two or more AGV because sharing a friendship simultaneously in the some time Cross road mouth and the conflict that causes, the most as shown in Figure 3.Conflicting for such, system typically uses waiting strategy to solve, i.e. The time window that AGV application low for priority uses is translated a period of time by system backward, and the AGV high until priority passes through Behind conflict intersection, then application uses this intersection, avoids deadlock and collision conflict with this.When multiple stage AGV arrives simultaneously During a certain intersection, first each AGV priority judged by system, then according to priority height sequencing AGV leads to Cross intersection sequencing.

Fig. 4 show employing waiting strategy and solves comparison diagram before and after the conflict of intersection time window, white in figure, black and Grey rectangle frame represents time window, the time window of AGV2 reservation application use and the conflict of two AGV that AGV1 registration uses respectively Time window.For avoiding AGV2 to clash with AGV1 during execution task, system can use waiting strategy being intersected by AGV2 On the i of crossing, the time window of application translates a rational time backward, when i.e. allowing AGV2 wait one section before entering intersection i Between, until intersection i is released, specifically as shown in figure b.When AGV2 passes through intersection i, system can be automatically AGV2 is disposed in time at the log-on message of intersection i, thus discharges this intersection resource, to facilitate other AGV Shens Please use.

Step S52: path conflict can be divided into again catches up with and surpasses conflict and conflicts in opposite directions, wherein, conflicts in opposite directions and can be divided into and can keep away Exempt from conflict and inevitable conflict.For may wait for avoiding conflict, system can use waiting strategy to solve；For inevitable punching Prominent, path planning strategy again can be used, this strategy includes local paths planning strategy and global path planning strategy；For catching up with Overshoot is dashed forward, and deceleration and waiting strategy can be used to solve, it is possible to according to actual needs, uses local paths planning strategy to feasible The conflict section in path carries out planning again and processes.

Further, in described step S52, in conflict refers to certain time period in opposite directions, same path is transported in opposite directions The conflict caused for contention path resources between the AGV of row, the most as shown in Figure 5.Catch up with and surpass conflict and refer to that two AGV are simultaneously one The speed of service of operation and the rear side AGV speed of service higher than front side AGV on paths, for contention path resources between them The conflict caused, the most as shown in Figure 7.

The conflict in opposite directions that Fig. 5 show in path conflict, in Fig. 5, (a) figure show and may wait for avoiding conflict, for this type of Conflict, system can use waiting strategy to solve, i.e. when the time window that AGV2 application low for priority uses is translated one section backward Between, after passing through intersection i until the AGV1 that priority is high, then application uses this intersection, avoids with this and AGV1 Deadlock and collision conflict occur, and the time window after adjustment is as shown in (b) figure in Fig. 6, and Fig. 6 show waiting strategy and solves path Conflict.In Fig. 5, (b) figure show inevitable conflict, conflicts for this type of, and maximally effective resolution policy is again to advise for AGV2 Draw new feasible path.Fig. 7 show and catches up with and surpasses conflict, from map analysis, between AGV1 and AGV2 only two conditions (i.e. The speed of service of AGV1 still keeps running in the same direction higher than AGV2 and AGV1 two cars before catching up with and surpassing AGV2) situation about simultaneously meeting Under, catch up with and surpass conflict and just can occur.For catching up with and surpassing conflict, if system does not take any control measure, same path is transported Rear-end collision must occur between two AGV of row, therefore, conflict for such, deceleration and waiting strategy can be used to solve, also Can use local paths planning strategy that the conflict section of feasible path carries out planning again and process according to actual needs.

Further, in described step S52, local paths planning strategy refers in not utilizing garage traffic network whole Road section information and in not changing the shortest feasible path on the premise of Lothrus apterus section, only on other roads adjacent with conflict section Duan Zhong, searches out other sections that can substitute conflict section, and this section can guarantee that and do not affecting on the shortest feasible path On the premise of the distribution of other times window, can guarantee that the AGV received an assignment arrives at the destination smoothly, the Transport Vehicle completing to specify is appointed Business.

Further, in described step S52, global path planning strategy refers to utilize whole sections in garage traffic network Information, the AGV for receiving an assignment cooks up the shortest feasible path again.

Further, in described step S52, the step that is embodied as of path planning includes again:

Step S521: system detects the inevitable conflict whether occurring in conflicting in opposite directions between many AGV；

Step S522: the section inevitable conflict occur is marked, and calls path search algorithm and again plan Path；

Step S523: to new search to the time window vector table of path optimizing be circulated renewal, until cooking up nothing Till the path optimizing that conflict is minimum with the time, algorithm search terminates.If algorithm is searched for (for avoiding program through successive ignition Endless loop occur, program cyclic search number of times is provided with maximum restriction) still cannot find Lothrus apterus path optimizing, then algorithm search Process terminates, and this task is loaded in task sequence table, and waiting system task scheduling next time is distributed.

Step S6: utilize parking system path planning algorithm based on dynamic time windows to plan Lothrus apterus optimum road for AGV Footpath, concrete steps include:

Assume that system has n platform AGV, current distribution task to have m item, and m item task need to assign m platform AGV to complete.For system In AGV, the AGV received an assignment, Transport Vehicle task, priority policy process after Transport Vehicle task and the starting point of task and Impact point can represent with set N, set Q, set M, set A, set S and set R respectively, i.e. N={n₁,n₂,n₃,…,n_n, Q ={ q₁,q₂,q₃,…,q_m, M={m₁,m₂,m₃,…,m_m, A={a₁,a₂,a₃,...,a_m, S={s₁,s₂,s₃,…,s_m, R ={ r₁,r₂,r₃,…,r_m}。

Step S61: each parameter of initialization algorithm, sets up set N, set Q, set M, set A, set S and set R, point Not Yong Yu AGV in storage system, the AGV received an assignment, the instruction of Transport Vehicle task requests, priority policy process after access Car task and the starting point of task and impact point；

Step S62: the Transport Vehicle request instruction in set M is loaded in task sequence set A, and according to priority of task Level height order is its sequence, the arrangement set A={a after process₁,a₂,a₃,...,a_m, a₁,a₂,a₃,...,a_mRepresent according to Task ranking after priority policy process, wherein, task a₁Priority is the highest, task a_mPriority is minimum；

Step S63: judge AGV duty according to AGV current state information, such as idle condition and be carrying out task shape State, state can represent with 0 and 1.If available free AGV exists in working environment, then system can be by preferential in task sequence set A Transport Vehicle task a that level is the highest₁Distribute to numbered q₁(q₁For the set the highest AGV of Q medium priority) idle AGV, and Know AGV start position s₁, aiming spot r₁And on the premise of the information such as working environment, call dijkstra's algorithm, for numbered q₁AGV cook up a shortest path optimizing of distance (this path can with orderly section gather λ₁Represent, i.e. λ₁={ e_j,e_k, e_l,…,e_q, e_j,e_k,e_l,…,e_q∈E₁, wherein, E₁Represent all feasible sections set in path optimizing), then go to step S64；Otherwise, then task scheduling stops；

Step S64: calculate AGV (numbered q₁) time of sailing in each section and roll the time away from path optimizing, according to Step S41 initializes each section time window vector table, by repeatedly cyclic search, the time window of the shortest i.e. available feasible path Distribution, usable setRepresent, then go to step S65；

Step S65: to suboptimum task a₂Path plan, go to step S63, it is judged that no available free AGV in environment Can call, if having, then system is first by numbered q₂Idle AGV distribute to suboptimum task a₂, then calling dijkstra's algorithm is Task a₂Cook up a path optimizing；Otherwise, then the scheduling to this task is suspended, then according to task sequence tabular order sequence is successively Other tasks are scheduling；If nothing, then system stops dispatching follow-up work；

Step S66: calculate AGV (numbered q₂) time of sailing in each section and rolling away from the time on sub-optimal path, update Each section time window vector table, then judges whether time window vector table exists overlap.If time window is non-overlapping, then suboptimum is appointed Business a₂Path planning process terminate；Otherwise, then need to detect suboptimum task a₂Path optimizing exist which kind of conflict, and according to Conflict type difference, select suitable conflict-solving strategy, as intersection conflict and in opposite directions conflict in avoid Conflict, can use waiting strategy to solve；For the inevitable conflict in conflict in opposite directions, path planning strategy solution again can be used Certainly；Deceleration and waiting strategy can be used to solve for catching up with and surpassing conflict, it is possible to according to actual needs, to use local paths planning plan The slightly conflict section to feasible path carries out the process of planning again；

Step S67: suboptimum task a₂Path planning terminate after, go to step S63, then according to task sequence tabular order sequence Repeat aforesaid operations, be sequentially completed the path planning of other tasks.

Further, in described step S63, system is the letter uploaded in real time with AGV self to the judgement of AGV status information Breath is as foundation, and the idle condition of AGV represents with 0, is carrying out task status and represents with 1.

Further, in described step S63, before the shortest feasible path of the mission planning that system is distributed, need to first judge Whether system exists idle AGV.If there is idle AGV, system just can be to mission planning feasible path；Otherwise, system cannot For the shortest feasible path of mission planning.

Further, in described step S63, feasible path the shortest to task planning with to accepting the shortest of this task AGV Feasible path planning is the same.

Further, in described step S65, suspended task is dispatched in two kinds of situation, respectively: 1. system is without idle AGV Schedulable uses, then system can stop dispatching follow-up work；2. in system, available free AGV exists, but system cannot be for this Lothrus apterus feasible path is cooked up in business.Now, system only suspends the scheduling to this task, and it has no effect on other tasks Scheduling.

Further, in described step S66, update each section time window vector table, it is judged that each section time window window vector Whether table exists overlapping phenomenon can operate according to step S42 and step S5.

Further, in described step S66, the time window arrangement of the shortest feasible path of suboptimum task can be counted according to step S4 Obtain.

Further, in described step S66, time window overlap problem, the strategy solution that can provide according to step S5.

The above is the exemplary description combining accompanying drawing to better embodiment of the present invention, and the present invention implements Not limited by aforesaid way, any those skilled in the art without departing from the spirit and scope of the present invention, may be by Technical solution of the present invention is made possible variation and amendment by the method for the disclosure above and technology contents, therefore, every without departing from The content of technical solution of the present invention, any simple modification above example made according to the technical spirit of the present invention, equivalent Change and modification, belong to the protection domain of technical solution of the present invention.

Claims (10)

1. a parking system paths planning method based on dynamic time windows, it is characterised in that comprise the steps:

Step S1: use topological approach to create the working environment model of AGV in intelligent garage；

Step S2: according to different evaluation standard, respectively every AGV and each Transport Vehicle task setting priority；

Step S3: using dijkstra's algorithm is that the AGV received an assignment plans the shortest feasible path；

Step S4: arrangement feasible path time window；

Step S5: different according to conflict type, design conflicts resolution policy；

Step S6: utilize parking system path planning algorithm based on dynamic time windows to plan Lothrus apterus optimal path for AGV.

2. the method for claim 1, it is characterised in that in described step S1, uses topological approach to create in intelligent garage The working environment model of AGV, concrete steps include:

Step S11: the traffic network in environmental model and AGV are handled as follows: 1. AGV runs track is the two-way mould of single track Formula, and width is only capable of accommodating an AGV；2. the AGV in system can only accept a Transport Vehicle within the same time period and appoints Business, during its execution task, it is invalid that other tasks of system distribution are then considered as；3. for avoiding colliding thing with other AGV Therefore, the AGV being required to be execution task sets a safety traffic region, and this safety zone can be by AGV car body physical dimension, operation speed Degree and operation track physical dimension determine；4. within certain moment or certain time period, the arbitrary intersection in road network and Arbitrary running section the most only allows an AGV to use；

Step S12: photographic head, radar sensor, ultrasonic sensor and the infrared ray sensor etc. that utilize AGV to carry gather AGV running environment information, above-mentioned information includes the initial parking stall of AGV, target parking stall, barrier and AGV position to be charged etc.；

Step S13: create AGV at intelligent garage as modeling data, employing topological approach using the environmental information that aforesaid operations gathers In working environment model.

3. the method for claim 1, it is characterised in that in described step S2, according to different evaluation standard, is respectively every Platform AGV and each Transport Vehicle task setting priority, particular content includes:

Step S21: for the priority of AGV in system, then determined by car number size, and AGV priority height order with Car number size becomes negative correlation；

Step S22: for the priority of Transport Vehicle task in system, then by task loading sequence, the task order of importance and emergency and distance The evaluation criterion such as the shortest comprehensively determines；

Step S23: when occur intersection conflict time, for AGV by conflict intersection sequencing problem, then by AGV current priority and the shortest priority of distance comprehensively determine；

Step S24: system it further provides for, the priority of the AGV being carrying out Transport Vehicle task is higher than the priority of idle AGV；? During AGV execution task, ground control system is that the new Transport Vehicle task of its distribution is considered invalid；

In described step S23, when intersection conflict occurs, for the AGV sequencing problem by intersection of conflicting, Situation about then comprehensively being determined by AGV current priority and the shortest priority of distance includes:

Step 231: when two AGV arrive same intersection simultaneously, first AGV priority judged, then by system Order according to the priority, arranges two AGV sequencing by intersection；When the AGV that priority is high passes through to hand over After cross road mouth and the AGV low with priority keep certain safe distance, the AGV that system meeting call priority is low continues executing with and appoints Business；

Step 232: when two AGV are one in front and one in back to arrive intersection, but when both meetings conflict occur in intersection, now System on the basis of judging AGV priority, also will according to the length of two AGV to intersection distance determine each other by The sequencing at crossing；

Step 233: when two AGV that priority is identical arrive intersection simultaneously, system can be intersected according to two AGV distances The distance at crossing determines its sequencing by intersection.

4. the method for claim 1, it is characterised in that in described step S3, uses dijkstra's algorithm for accepting to appoint Business AGV plan the shortest feasible path it is critical that its must according in step S2 priority height order carry out；For Task m any one of system_iFunction may be defined as:

m_i(t)=(s_i,d_i,λ_i(t),P_i(t),q_i)

In formula, i represents mission number；m_iT () represents the task of t system distribution；s_iExpression task m_iStarting point, d_iRepresent and appoint Business m_iTerminal, λ_iExpression task m_iThe set in a series of orderly section of process, P_iT () represents task m_iPriority, q_i Represent execution task m_iAGV；After many AGV path planning terminates, the above-mentioned parameter of each task typically immobilizes, only In case of a collision, the AGV that some priority is low just needs dynamically to change its running route, avoids performing to appoint with this Business AGV between collide, deadlock conflict and strengthen AGV flexibility.

5. the method for claim 1, it is characterised in that in described step S4, feasible path time window of arranging, specifically walk Suddenly include:

Step S41: time window initializes；After the shortest feasible path determines, under ideal conditions, for the AGV arrangement received an assignment Go out feasible path time window；By task m in step S3_iThe shortest feasible path λ found out_i, it is by a series of operation sections group Become, available orderly section set expression, i.e. λ_i={ e_j,e_k,e_l,…,e_q, e_j,e_k,e_l,…,e_q∈ E, wherein, E represents road network In the set in all feasible sections, e_k(k ∈ 1,2,3 ...., q) represent a certain section in the shortest feasible path；

Task m_iAt section e_kOn time window function may be defined as:

T_w,ik=(q_i,m_i,r,t_in,k,t_out,k)

In formula, r represents section e_kAt feasible path λ_iOn position；t_in,kRepresent vehicle q_iSail section e into_kInitial time； t_out,kRepresent vehicle q_iLeave section e_kTime；

For section e_kTime window, can be calculated by following formula:

t_out,k=t_in,k+ω_i,k

In formula, ω_i,kRepresent that AGV is by section e_kThe time spent, can be calculated by following formula:

${\mathrm{\ω}}_{i,k}=\frac{l_{i,k}}{v}$

In formula, l_i,kRepresent section e_kLength, v represents the speed of service of AGV；

In actual applications, due to the section e the most in order of feasible path_kNeed to be used by AGV timesharing, therefore, orderly section e_k Also it is made up of a series of time windows, available ordered vector Represent；At ordered vector In, vector dimension is identical with Transport Vehicle task quantity, can change over and change；If task m_iDo not use in certain moment Section e_k, then can be time of sailing into the t in this section_in,kWith roll time t away from_out,kIt is both configured to 0；Further, since task m_i? Short feasible path is made up of a series of orderly sections, and every orderly section correspond to a time window, therefore, and task m_i It is also believed to be made up of a series of time windows, usable setRepresent；

According to equation given in step S41, it can be task m_iThe shortest feasible path λ_iArrange out such as set D_iThe shown time Window is distributed；

Step S42: time window updates；Arrange out behind a time window path ideally according to step S41, then check Between different task, whether the time window of feasible path exists overlapping phenomenon；

If non-overlapping phenomenon, then task m_iPath planning process terminate；If current task m_iBe current system medium priority During high scheduler task, then feasible path time window step S41 planned is as task m_iFinal time window, it is not necessary to again Update；

If there being lap, then an at least section on explanation current task and the shortest feasible path that goes out of other mission plannings Use simultaneously；For this kind of phenomenon, then need system according to conflict type, conflict Robot dodge strategy reasonable in design；

Step S43: the arrangement of time window；Lothrus apterus time window corresponding for each for feasible path section is arranged in a certain order Row, i.e. complete the time window arrangement of feasible path；If a certain section e_kThere is the time window of multiple task, then the task that is newly added exists Section e_kOn the entry time of time window must be fulfilled for condition: 1. sail into time in this section have to be larger than or equal to AGV from The time departure in a upper section；2. the length of the free time window in this section be should be greater than or be spent by this section equal to AGV The time taken；

Described time window refers to that the AGV performing Transport Vehicle task leaves certain intersection or certain section from initially entering to The time that whole process is spent, its Main Function is the intersection taken AGV or running section is marked.

6. the method for claim 1, it is characterised in that according to step S3 and step S4, searched for by loop iteration, can It is followed successively by the AGV received an assignment and cooks up Lothrus apterus the shortest feasible path time window.

7. the method for claim 1, it is characterised in that in described step S4, ground control system can real-time reception AGV The information such as position, speed and the running status uploaded during exercising, and according to these information, it is judged that AGV the most from Open certain section or certain intersection, drive towards next section or intersection；If AGV have been moved off certain section or certain Intersection, then need to delete the information that AGV registers in the table from time window vector table, thus discharge this section maybe this intersection Crossing resource, for other AGV；

In described step S4, to avoid within the time period that this AGV takies, used by other AGV and cause deadlock or collision punching Prominent.

8. the method for claim 1, it is characterised in that in described step S5, different according to conflict type, design conflicts Resolution policy, particular content includes:

The conflict occurred in many AGV Transport Vehicle path planning typically has two kinds, i.e. intersection conflict and path conflict,

Step S51: intersection conflict refers to be carved with two or more AGV because sharing a crossroad simultaneously in the some time Mouthful and the conflict that causes；Conflicting for such, system typically uses waiting strategy to solve, i.e. system is by AGV low for priority The time window that application uses translates a period of time backward, after passing through conflict intersection until the AGV that priority is high, then Shen This intersection be please use, deadlock and collision conflict avoided with this；When multiple stage AGV arrives a certain intersection simultaneously, system First each AGV priority is judged, then according to priority height sequencing AGV is by intersection sequencing；

Step S52: path conflict can be divided into again and conflicts in opposite directions and catch up with and surpass conflict, wherein, conflicts in opposite directions and can be divided into and can avoid punching Prominent and inevitable conflict；For may wait for avoiding conflict, system can use waiting strategy to solve；For inevitable conflict, Can use path planning strategy again, this strategy includes local paths planning strategy and global path planning strategy；For catching up with and surpassing Conflict, can use deceleration and waiting strategy to solve, it is possible to according to actual needs, uses local paths planning strategy to can walking along the street The conflict section in footpath carries out planning again and processes；

In described step S52, local paths planning strategy refer in not utilizing garage traffic network whole road section informations and Do not change in the shortest feasible path on the premise of Lothrus apterus section, only in other sections adjacent with conflict section, search out Article one, other sections in conflict section can be substituted, and this section can guarantee that and divides not affecting other times window on the shortest feasible path On the premise of cloth, can guarantee that the AGV received an assignment arrives at the destination smoothly, complete the Transport Vehicle task specified；

In described step S52, global path planning strategy refers to utilize whole road section informations in garage traffic network, again for connecing The shortest feasible path is cooked up by the AGV of task；

In described step S52, the step that is embodied as of path planning includes again:

Step S521: system detects the inevitable conflict whether occurring in conflicting in opposite directions between many AGV；

Step S522: the section inevitable conflict occur is marked, and calls path search algorithm path planning again；

Step S523: to new search to the time window vector table of path optimizing be circulated renewal, until cooking up Lothrus apterus Till the path optimizing minimum with the time, algorithm search terminates；If algorithm still cannot find without punching through successive ignition search Prominent path optimizing, then algorithm search process terminates, and this task is loaded in task sequence table, and under waiting system, subtask is adjusted Degree distribution；

In described step S52, in conflict refers to certain time period in opposite directions, for striving between the AGV run in opposite directions on same path The conflict taking path resources by force and cause；Catch up with and surpass conflict and refer to that two AGV run and the operation of rear side AGV the most on one path The speed speed of service higher than front side AGV, the conflict caused for contention path resources between them.

9. the method for claim 1, it is characterised in that in described step S6, utilizes based on dynamic time windows parking System path planning algorithm is that AGV plans Lothrus apterus optimal path, and concrete steps include:

Step S61: each parameter of initialization algorithm, sets up set N, set Q, set M, set A, set S and set R, uses respectively Transport Vehicle after the instruction of AGV in storage system, the AGV received an assignment, Transport Vehicle task requests, priority policy process is appointed Business and the starting point of task and impact point；

Step S62: the Transport Vehicle request instruction in set M is loaded in task sequence set A, and high according to task priority Low order is its sequence；

Step S63: judge AGV duty according to AGV current state information, such as idle condition and be carrying out task status, State can represent with 0 and 1；If available free AGV exists in working environment, then system can be by task sequence set A medium priority The highest Transport Vehicle task distributes to the idle AGV that numbering is minimum, and in known AGV start position, aiming spot and work On the premise of the information such as environment, call dijkstra's algorithm, cook up a shortest path optimizing of distance for AGV, then turn To step S64；Otherwise, then task scheduling stops；

Step S64: calculate AGV time of sailing in each section and roll the time away from path optimizing, initializes each according to step S41 Section time window vector table, by repeatedly cyclic search, the time window distribution of the shortest i.e. available feasible path, then goes to step Rapid S65；

Step S65: the path of suboptimum task is planned, goes to step S63, it is judged that in environment, no available free AGV can call, If having, then numbering is first only second to the highest idle AGV of priority and distributes to suboptimum task by system, then calls Dijkstra and calculates Method is that suboptimum mission planning goes out a path optimizing；Otherwise, then suspend the scheduling to this task, then according to task sequence tabular order Other tasks are scheduling by sequence successively；If nothing, then system stops dispatching follow-up work；

Step S66: calculate AGV time of sailing in each section and roll the time away from sub-optimal path, update each section time window to Scale, then judges whether time window vector table exists overlap；If time window is non-overlapping, then the path planning mistake of suboptimum task Journey terminates；Otherwise, then need which kind of conflict the path optimizing detecting suboptimum task exists, and according to the difference of conflict type, choosing Use suitable conflict-solving strategy, as intersection conflict and in opposite directions conflict in the conflict avoided, wait plan can be used Slightly solve；For the inevitable conflict in conflict in opposite directions, path planning strategy can be used again to solve；Can for catching up with and surpassing conflict Use deceleration and waiting strategy to solve, it is possible to according to actual needs, use local paths planning strategy that feasible path is rushed Prominent section carries out planning again and processes；

Step S67: after the path planning of suboptimum task terminates, go to step S63, then according on task sequence tabular order sequence repeats State operation, be sequentially completed the path planning of other tasks.

10. method as claimed in claim 9, it is characterised in that in described step S63, the system judgement to AGV status information Be the information uploaded in real time using AGV self as foundation, and the idle condition of AGV represents with 0, is carrying out task status with 1 Represent；

In described step S63, before the shortest feasible path of the mission planning that system is distributed, need to first judge whether system exists Idle AGV；If there is idle AGV, system just can be to mission planning feasible path；Otherwise, system cannot be the shortest for mission planning Feasible path；

In described step S63, feasible path the shortest to task planning with planning the shortest feasible path accepting this task AGV is The same；

In described step S65, suspended task is dispatched in two kinds of situation, respectively: 1. system uses, then without idle AGV schedulable System can stop dispatching follow-up work；2. in system, available free AGV exists, but system cannot go out Lothrus apterus for this mission planning Feasible path；Now, system only suspends the scheduling to this task, and it has no effect on the scheduling to other tasks；

In described step S66, update each section time window vector table, it is judged that whether each section time window window vector table exists overlap Phenomenon can operate according to step S42 and step S5；

In described step S66, the time window arrangement of the shortest feasible path of suboptimum task can be calculated according to step S4；

In described step S66, time window overlap problem, the strategy solution that can provide according to step S5.